Tagging support apparatus and method

Abstract

According to an embodiment, a tagging support apparatus includes a first acquirer, an estimator, a first storage, a second acquirer, and a presenter. The first acquirer acquires a spoken sentence concerning an utterance of a user. The estimator estimates an utterance intention of the spoken sentence to obtain one or more intention candidates of the utterance intention. The first storage stores an intention system having a hierarchical structure of intentions used in a dialog system. The second acquirer acquires, based on the intention candidates, part of the intention system as one or more hierarchical intention candidates. The presenter presents the hierarchical intention candidates.

The invention claimed is: 
     
       1. A tagging support apparatus comprising:
 a first acquirer configured to acquire a spoken sentence concerning an utterance of a user; 
 an estimator configured to estimate an utterance intention of the spoken sentence to obtain one or more intention candidates of the utterance intention; 
 a first storage configured to store an intention system having a hierarchical structure of intentions used in a dialogue system; 
 a second acquirer configured to acquire, based on the intention candidates, part of the intention system as one or more hierarchical intention candidates; 
 a converter configured to convert a hierarchical structure of the hierarchical intention candidates such that an intention whose number of appearances as a subordinate concept is not less than a threshold changes to a superordinate concept; and 
 a presenter configured to present the hierarchical intention candidates whose hierarchical structure is converted. 
 
     
     
       2. The apparatus according to  claim 1 , further comprising a selector configured to select a hierarchical intention candidate as a selected intention from the hierarchical intention candidates in accordance with an instruction from one of the user and a worker. 
     
     
       3. The apparatus according to  claim 2 , further comprising a second storage configured to store the spoken sentence and the selected intention in association with each other. 
     
     
       4. The apparatus according to  claim 1 , wherein the second acquirer is configured to acquire at least one of (a) an intention corresponding to a superordinate concept of at least one of the intention candidates, and (b) an intention corresponding to a subordinate concept of at least one of the intention candidates. 
     
     
       5. The apparatus according to  claim 1 , wherein the second acquirer is configured to acquire the hierarchical intention candidates from the intention system whose hierarchical structure is converted. 
     
     
       6. The apparatus according to  claim 1 , wherein the converter is configured to convert the hierarchical structure such that an intention of a clause out of at least one clause of the spoken sentence, whose certainty of representing the intention is not less than the threshold, changes to a superordinate concept. 
     
     
       7. The apparatus according to  claim 1 , wherein the estimator is configured to perform character string matching between the spoken sentence and a descriptive text of an intention tag to obtain the intention candidates. 
     
     
       8. The apparatus according to  claim 1 , wherein the estimator is configured to obtain N-best of estimated intentions as the intention candidates. 
     
     
       9. The apparatus according to  claim 1 , wherein the second acquirer is configured to acquire one or more parts of the intention system as the hierarchical intention candidates while holding a structure of the intention system. 
     
     
       10. The apparatus according to  claim 3 , wherein the presenter is configured to present an uppermost layer of the hierarchical intention candidates, and to expand and present a lower layer of the hierarchical intention candidates in accordance with the instruction of the worker. 
     
     
       11. A tagging support method comprising:
 acquiring a spoken sentence concerning an utterance of a user; 
 estimating an utterance intention of the spoken sentence to obtain one or more intention candidates of the utterance intention; 
 storing an intention system having a hierarchical structure of intentions used in a dialogue system; 
 acquiring, based on the intention candidates, part of the intention system as one or more hierarchical intention candidates; 
 converting a hierarchical structure of the hierarchical intention candidates such that an intention whose number of appearances as a subordinate concept is not less than a threshold changes to a superordinate concept; and 
 presenting the hierarchical intention candidates whose hierarchical structure is converted. 
 
     
     
       12. A non-transitory computer readable storage medium storing instructions of a computer program which when executed by a computer results in performance of steps comprising:
 acquiring a spoken sentence concerning an utterance of a user; 
 estimating an utterance intention of the spoken sentence to obtain one or more intention candidates of the utterance intention; 
 storing an intention system having a hierarchical structure of intentions used in a dialogue system; 
 acquiring, based on the intention candidates, part of the intention system as one or more hierarchical intention candidates; 
 converting a hierarchical structure of the hierarchical intention candidates such that an intention whose number of appearances as a subordinate concept is not less than a threshold changes to a superordinate concept; and 
 presenting the hierarchical intention candidates whose hierarchical structure is converted.